Face Generation¶

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data¶

You'll be using two datasets in this project:

MNIST
CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

#data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Found mnist Data
Found celeba Data

Explore the Data¶

MNIST¶

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 50

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

/usr/local/lib/python3.5/site-packages/matplotlib/font_manager.py:280: UserWarning: Matplotlib is building the font cache using fc-list. This may take a moment.
  'Matplotlib is building the font cache using fc-list. '

<matplotlib.image.AxesImage at 0x7f092fb158d0>

CelebA¶

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 50

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7f092fa39860>

Preprocess the Data¶

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network¶

You'll build the components necessary to build a GANs by implementing the following functions below:

model_inputs
discriminator
generator
model_loss
model_opt
train

Check the Version of TensorFlow and Access to GPU¶

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.1.0
Default GPU Device: /gpu:0

Input¶

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
Z input placeholder with rank 2 using z_dim.
Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    inputs_real = tf.placeholder(tf.float32, (None, image_width, image_height, image_channels), name='inputs_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='inputs_z')
    learning_rate = tf.placeholder(tf.float32, name='learning_rate')

    return (inputs_real, inputs_z, learning_rate)


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

Tests Passed

Discriminator¶

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the discriminator, tensor logits of the discriminator).

def discriminator(images, reuse=False, alpha=0.2, keep_prob=1):
    """
    Create the discriminator network
    :param images: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    with (tf.variable_scope('discriminator', reuse=reuse)):
        # Input layer is 28x28x3
        x1 = tf.layers.conv2d(images, 64, 5, strides=2, padding='same')
        relu1 = tf.maximum(alpha * x1, x1)
        # 14x14x64
        
        x2 = tf.layers.conv2d(relu1, 128, 5, strides=2, padding='same')
        drop2 = tf.nn.dropout(x2, keep_prob)
        bn2 = tf.layers.batch_normalization(drop2, training=True)
        relu2 = tf.maximum(alpha * bn2, bn2)
        # 7x7x128
        
        x3 = tf.layers.conv2d(relu2, 256, 5, strides=1, padding='same')
        drop3 = tf.nn.dropout(x3, keep_prob)
        bn3 = tf.layers.batch_normalization(drop3, training=True)
        relu3 = tf.maximum(alpha * bn3, bn3)
        # 7x7x256

        # Flatten it
        flat = tf.reshape(relu3, (-1, 7*7*256))
        logits = tf.layers.dense(flat, 1)
        out = tf.sigmoid(logits)
        
        return (out, logits)

"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator¶

Implement generator to generate an image using z. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, is_train=True, alpha=0.2):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    with (tf.variable_scope('generator', reuse=not is_train)):
        # First fully connected layer
        x1 = tf.layers.dense(z, 7*7*512)
        # Reshape it to start the convolutional stack
        x1 = tf.reshape(x1, (-1, 7, 7, 512))
        x1 = tf.layers.batch_normalization(x1, training=is_train)
        x1 = tf.maximum(alpha * x1, x1)
        # 7x7x512 now
        
        x2 = tf.layers.conv2d_transpose(x1, 256, 5, strides=2, padding='same')
        x2 = tf.layers.batch_normalization(x2, training=is_train)
        x2 = tf.maximum(alpha * x2, x2)
        # 14x14x256 now
        
        x3 = tf.layers.conv2d_transpose(x2, 128, 5, strides=2, padding='same')
        x3 = tf.layers.batch_normalization(x3, training=is_train)
        x3 = tf.maximum(alpha * x3, x3)
        # 28x28x128 now
        
        # Output layer
        logits = tf.layers.conv2d_transpose(x3, out_channel_dim, 5, strides=1, padding='same')
        # 28x28x3 now
        
        out = tf.tanh(logits)
    
    return out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss¶

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

discriminator(images, reuse=False)
generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim, alpha=0.2):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    g_model = generator(input_z, out_channel_dim, alpha=alpha)
    d_model_real, d_logits_real = discriminator(input_real, alpha=alpha)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True, alpha=alpha)

    d_loss_real = tf.reduce_mean(
                        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels=tf.ones_like(d_model_real)))
    
    d_loss_fake = tf.reduce_mean(
                        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
    
    g_loss = tf.reduce_mean(
                        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    d_loss = d_loss_real + d_loss_fake
    
    return d_loss, g_loss



"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization¶

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train \
                            .AdamOptimizer(learning_rate, beta1=beta1) \
                            .minimize(d_loss, var_list=d_vars)
                
        g_train_opt = tf.train \
                            .AdamOptimizer(learning_rate, beta1=beta1) \
                            .minimize(g_loss, var_list=g_vars)

    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training¶

Show Output¶

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train¶

Implement train to build and train the GANs. Use the following functions you implemented:

model_inputs(image_width, image_height, image_channels, z_dim)
model_loss(input_real, input_z, out_channel_dim)
model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    #saver = tf.train.Saver()
    
    image_width = data_shape[1]
    image_height = data_shape[2]
    image_channels = data_shape[3]
    
    input_real, input_z, lr = model_inputs(image_width, image_height, image_channels, z_dim)
    d_loss, g_loss = model_loss(input_real, input_z, image_channels, alpha=0.2)
    d_opt, g_opt = model_opt(d_loss, g_loss, learning_rate, beta1)
    
    steps = 0
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
                steps += 1
                
                # Sample random noise for G
                batch_z = np.random.uniform(-1, 1, size=(batch_size, z_dim))
                
                # Run optimizers
                _ = sess.run(d_opt, feed_dict={input_real: batch_images, input_z: batch_z})
                _ = sess.run(g_opt, feed_dict={input_z: batch_z, input_real: batch_images})
                
                if steps % 10 == 0:
                    # At the end of each epoch, get the losses and print them out
                    train_loss_d = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    train_loss_g = g_loss.eval({input_z: batch_z, input_real: batch_images})
                    
                    print("Epoch {}/{}...".format(epoch_i+1, epoch_count),
                          "Discriminator Loss: {:.4f}...".format(train_loss_d),
                          "Generator Loss: {:.4f}".format(train_loss_g))
                    
                if steps % 100 == 0:
                    show_generator_output(sess, 16, input_z, image_channels, data_image_mode)
                    
        #saver.save(sess, './checkpoints/generator.ckpt')

MNIST¶

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 64
z_dim = 100
learning_rate = 0.0002
beta1 = 0.5

"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)

Epoch 1/2... Discriminator Loss: 1.0511... Generator Loss: 0.6482
Epoch 1/2... Discriminator Loss: 0.0580... Generator Loss: 4.1681
Epoch 1/2... Discriminator Loss: 5.3844... Generator Loss: 0.0067
Epoch 1/2... Discriminator Loss: 0.1765... Generator Loss: 3.4673
Epoch 1/2... Discriminator Loss: 3.6017... Generator Loss: 0.0783
Epoch 1/2... Discriminator Loss: 0.8012... Generator Loss: 1.2852
Epoch 1/2... Discriminator Loss: 1.1833... Generator Loss: 0.9899
Epoch 1/2... Discriminator Loss: 0.8326... Generator Loss: 1.0630
Epoch 1/2... Discriminator Loss: 0.8427... Generator Loss: 2.0986
Epoch 1/2... Discriminator Loss: 0.9614... Generator Loss: 0.6862

Epoch 1/2... Discriminator Loss: 0.4847... Generator Loss: 1.3224
Epoch 1/2... Discriminator Loss: 0.3575... Generator Loss: 2.0151
Epoch 1/2... Discriminator Loss: 0.2232... Generator Loss: 2.2824
Epoch 1/2... Discriminator Loss: 0.1606... Generator Loss: 2.4390
Epoch 1/2... Discriminator Loss: 0.1189... Generator Loss: 2.5581
Epoch 1/2... Discriminator Loss: 0.0543... Generator Loss: 3.9214
Epoch 1/2... Discriminator Loss: 0.0408... Generator Loss: 3.8089
Epoch 1/2... Discriminator Loss: 0.0617... Generator Loss: 3.3983
Epoch 1/2... Discriminator Loss: 0.0402... Generator Loss: 3.6362
Epoch 1/2... Discriminator Loss: 0.0394... Generator Loss: 3.5887

Epoch 1/2... Discriminator Loss: 0.0321... Generator Loss: 3.9815
Epoch 1/2... Discriminator Loss: 0.0766... Generator Loss: 2.9582
Epoch 1/2... Discriminator Loss: 0.0772... Generator Loss: 3.2861
Epoch 1/2... Discriminator Loss: 0.0607... Generator Loss: 3.1192
Epoch 1/2... Discriminator Loss: 0.0246... Generator Loss: 4.1905
Epoch 1/2... Discriminator Loss: 0.0433... Generator Loss: 3.8217
Epoch 1/2... Discriminator Loss: 0.0744... Generator Loss: 2.8679
Epoch 1/2... Discriminator Loss: 2.2185... Generator Loss: 0.3431
Epoch 1/2... Discriminator Loss: 0.1546... Generator Loss: 3.0189
Epoch 1/2... Discriminator Loss: 0.1439... Generator Loss: 2.3280

Epoch 1/2... Discriminator Loss: 0.1034... Generator Loss: 2.9082
Epoch 1/2... Discriminator Loss: 0.0317... Generator Loss: 6.2053
Epoch 1/2... Discriminator Loss: 0.0196... Generator Loss: 5.9068
Epoch 1/2... Discriminator Loss: 0.0967... Generator Loss: 2.6962
Epoch 1/2... Discriminator Loss: 4.4948... Generator Loss: 12.2707
Epoch 1/2... Discriminator Loss: 0.2607... Generator Loss: 3.7203
Epoch 1/2... Discriminator Loss: 0.1830... Generator Loss: 4.1057
Epoch 1/2... Discriminator Loss: 1.1565... Generator Loss: 0.5118
Epoch 1/2... Discriminator Loss: 0.1333... Generator Loss: 11.1048
Epoch 1/2... Discriminator Loss: 0.0366... Generator Loss: 8.9235

Epoch 1/2... Discriminator Loss: 0.0201... Generator Loss: 15.0410
Epoch 1/2... Discriminator Loss: 1.1524... Generator Loss: 0.6501
Epoch 1/2... Discriminator Loss: 0.5427... Generator Loss: 1.5165
Epoch 1/2... Discriminator Loss: 0.1928... Generator Loss: 3.5568
Epoch 1/2... Discriminator Loss: 1.3003... Generator Loss: 0.5384
Epoch 1/2... Discriminator Loss: 0.0230... Generator Loss: 5.8277
Epoch 1/2... Discriminator Loss: 0.0043... Generator Loss: 11.2324
Epoch 1/2... Discriminator Loss: 0.1172... Generator Loss: 3.6143
Epoch 1/2... Discriminator Loss: 0.2906... Generator Loss: 7.9551
Epoch 1/2... Discriminator Loss: 0.0602... Generator Loss: 7.3511

Epoch 1/2... Discriminator Loss: 0.6172... Generator Loss: 0.9735
Epoch 1/2... Discriminator Loss: 0.0266... Generator Loss: 8.0473
Epoch 1/2... Discriminator Loss: 0.2255... Generator Loss: 11.3277
Epoch 1/2... Discriminator Loss: 0.5699... Generator Loss: 1.7674
Epoch 1/2... Discriminator Loss: 0.0923... Generator Loss: 7.4856
Epoch 1/2... Discriminator Loss: 0.0326... Generator Loss: 6.6476
Epoch 1/2... Discriminator Loss: 0.5984... Generator Loss: 6.0027
Epoch 1/2... Discriminator Loss: 0.0212... Generator Loss: 6.0808
Epoch 1/2... Discriminator Loss: 1.1500... Generator Loss: 0.5593
Epoch 1/2... Discriminator Loss: 0.2744... Generator Loss: 2.2808

Epoch 1/2... Discriminator Loss: 0.1494... Generator Loss: 3.0293
Epoch 1/2... Discriminator Loss: 0.1306... Generator Loss: 3.8457
Epoch 1/2... Discriminator Loss: 2.2580... Generator Loss: 8.7079
Epoch 1/2... Discriminator Loss: 1.2393... Generator Loss: 0.4766
Epoch 1/2... Discriminator Loss: 0.4668... Generator Loss: 1.4021
Epoch 1/2... Discriminator Loss: 0.2418... Generator Loss: 2.0790
Epoch 1/2... Discriminator Loss: 0.0351... Generator Loss: 6.5805
Epoch 1/2... Discriminator Loss: 0.2131... Generator Loss: 10.2824
Epoch 1/2... Discriminator Loss: 0.1146... Generator Loss: 6.0870
Epoch 1/2... Discriminator Loss: 1.5897... Generator Loss: 0.3181

Epoch 1/2... Discriminator Loss: 0.0344... Generator Loss: 8.0379
Epoch 1/2... Discriminator Loss: 0.0193... Generator Loss: 9.8849
Epoch 1/2... Discriminator Loss: 0.2321... Generator Loss: 9.6907
Epoch 1/2... Discriminator Loss: 0.0264... Generator Loss: 9.0274
Epoch 1/2... Discriminator Loss: 0.0395... Generator Loss: 5.5007
Epoch 1/2... Discriminator Loss: 0.1146... Generator Loss: 5.4601
Epoch 1/2... Discriminator Loss: 0.3030... Generator Loss: 1.9188
Epoch 1/2... Discriminator Loss: 0.1144... Generator Loss: 3.1901
Epoch 1/2... Discriminator Loss: 0.0301... Generator Loss: 6.1403
Epoch 1/2... Discriminator Loss: 0.0230... Generator Loss: 6.5453

Epoch 1/2... Discriminator Loss: 0.2181... Generator Loss: 6.5260
Epoch 1/2... Discriminator Loss: 0.2245... Generator Loss: 2.7372
Epoch 1/2... Discriminator Loss: 0.0417... Generator Loss: 4.0542
Epoch 1/2... Discriminator Loss: 0.0520... Generator Loss: 8.3141
Epoch 1/2... Discriminator Loss: 0.0832... Generator Loss: 3.1691
Epoch 1/2... Discriminator Loss: 0.2420... Generator Loss: 1.8937
Epoch 1/2... Discriminator Loss: 0.1307... Generator Loss: 2.6792
Epoch 1/2... Discriminator Loss: 0.1966... Generator Loss: 2.3472
Epoch 1/2... Discriminator Loss: 0.0590... Generator Loss: 3.3635
Epoch 1/2... Discriminator Loss: 0.5623... Generator Loss: 1.0826

Epoch 1/2... Discriminator Loss: 0.0822... Generator Loss: 5.7030
Epoch 1/2... Discriminator Loss: 0.0669... Generator Loss: 6.2832
Epoch 1/2... Discriminator Loss: 0.0445... Generator Loss: 7.9947
Epoch 2/2... Discriminator Loss: 0.1476... Generator Loss: 2.3671
Epoch 2/2... Discriminator Loss: 0.0646... Generator Loss: 3.6153
Epoch 2/2... Discriminator Loss: 0.1325... Generator Loss: 2.5069
Epoch 2/2... Discriminator Loss: 2.0565... Generator Loss: 0.2049
Epoch 2/2... Discriminator Loss: 0.5681... Generator Loss: 4.0676
Epoch 2/2... Discriminator Loss: 0.3364... Generator Loss: 1.8587
Epoch 2/2... Discriminator Loss: 0.2286... Generator Loss: 2.0696

CelebA¶

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size = 64
z_dim = 100
learning_rate = 0.0002
beta1 = 0.5


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 8

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

Epoch 1/8... Discriminator Loss: 0.6728... Generator Loss: 1.5186
Epoch 1/8... Discriminator Loss: 1.1272... Generator Loss: 0.8372
Epoch 1/8... Discriminator Loss: 0.3232... Generator Loss: 2.3715
Epoch 1/8... Discriminator Loss: 0.1226... Generator Loss: 4.2754
Epoch 1/8... Discriminator Loss: 0.9746... Generator Loss: 0.7104
Epoch 1/8... Discriminator Loss: 0.3357... Generator Loss: 2.6196
Epoch 1/8... Discriminator Loss: 0.4234... Generator Loss: 1.6064
Epoch 1/8... Discriminator Loss: 0.5877... Generator Loss: 1.1269
Epoch 1/8... Discriminator Loss: 0.1138... Generator Loss: 4.6090
Epoch 1/8... Discriminator Loss: 0.0383... Generator Loss: 4.0158

Epoch 1/8... Discriminator Loss: 0.0431... Generator Loss: 5.1493
Epoch 1/8... Discriminator Loss: 0.0573... Generator Loss: 5.8116
Epoch 1/8... Discriminator Loss: 0.0197... Generator Loss: 6.1787
Epoch 1/8... Discriminator Loss: 0.1315... Generator Loss: 2.5443
Epoch 1/8... Discriminator Loss: 0.0390... Generator Loss: 4.0249
Epoch 1/8... Discriminator Loss: 1.9109... Generator Loss: 0.3166
Epoch 1/8... Discriminator Loss: 0.0496... Generator Loss: 5.9687
Epoch 1/8... Discriminator Loss: 0.0226... Generator Loss: 7.6569
Epoch 1/8... Discriminator Loss: 0.0425... Generator Loss: 4.4188
Epoch 1/8... Discriminator Loss: 0.0176... Generator Loss: 11.2877

Epoch 1/8... Discriminator Loss: 0.0046... Generator Loss: 10.1641
Epoch 1/8... Discriminator Loss: 0.0157... Generator Loss: 4.7534
Epoch 1/8... Discriminator Loss: 0.0167... Generator Loss: 8.4297
Epoch 1/8... Discriminator Loss: 0.0059... Generator Loss: 6.9933
Epoch 1/8... Discriminator Loss: 0.1426... Generator Loss: 2.6345
Epoch 1/8... Discriminator Loss: 0.1045... Generator Loss: 4.6590
Epoch 1/8... Discriminator Loss: 0.0116... Generator Loss: 8.6840
Epoch 1/8... Discriminator Loss: 3.2107... Generator Loss: 0.1940
Epoch 1/8... Discriminator Loss: 1.1724... Generator Loss: 2.4985
Epoch 1/8... Discriminator Loss: 0.1298... Generator Loss: 3.7857

Epoch 1/8... Discriminator Loss: 0.0436... Generator Loss: 5.1883
Epoch 1/8... Discriminator Loss: 0.3382... Generator Loss: 2.9410
Epoch 1/8... Discriminator Loss: 0.2581... Generator Loss: 3.7506
Epoch 1/8... Discriminator Loss: 0.7456... Generator Loss: 0.8706
Epoch 1/8... Discriminator Loss: 0.2229... Generator Loss: 2.7091
Epoch 1/8... Discriminator Loss: 0.0117... Generator Loss: 9.6059
Epoch 1/8... Discriminator Loss: 1.2423... Generator Loss: 0.4302
Epoch 1/8... Discriminator Loss: 0.0857... Generator Loss: 6.3817
Epoch 1/8... Discriminator Loss: 2.5465... Generator Loss: 0.3115
Epoch 1/8... Discriminator Loss: 2.7727... Generator Loss: 7.5936

Epoch 1/8... Discriminator Loss: 0.1478... Generator Loss: 2.7957
Epoch 1/8... Discriminator Loss: 0.1403... Generator Loss: 6.0365
Epoch 1/8... Discriminator Loss: 2.2179... Generator Loss: 0.1487
Epoch 1/8... Discriminator Loss: 0.1272... Generator Loss: 3.6513
Epoch 1/8... Discriminator Loss: 0.1594... Generator Loss: 5.3392
Epoch 1/8... Discriminator Loss: 0.2873... Generator Loss: 2.7121
Epoch 1/8... Discriminator Loss: 0.1165... Generator Loss: 3.3148
Epoch 1/8... Discriminator Loss: 0.5408... Generator Loss: 8.0897
Epoch 1/8... Discriminator Loss: 0.4946... Generator Loss: 11.5567
Epoch 1/8... Discriminator Loss: 0.1720... Generator Loss: 3.2248

Epoch 1/8... Discriminator Loss: 1.3815... Generator Loss: 5.5735
Epoch 1/8... Discriminator Loss: 0.1686... Generator Loss: 10.5814
Epoch 1/8... Discriminator Loss: 0.1419... Generator Loss: 3.8295
Epoch 1/8... Discriminator Loss: 0.1603... Generator Loss: 5.5645
Epoch 1/8... Discriminator Loss: 0.0745... Generator Loss: 5.1192
Epoch 1/8... Discriminator Loss: 0.2997... Generator Loss: 2.4982
Epoch 1/8... Discriminator Loss: 0.4996... Generator Loss: 1.5383
Epoch 1/8... Discriminator Loss: 0.0705... Generator Loss: 7.0556
Epoch 1/8... Discriminator Loss: 0.0278... Generator Loss: 8.2699
Epoch 1/8... Discriminator Loss: 0.1870... Generator Loss: 2.6217

Epoch 1/8... Discriminator Loss: 4.7258... Generator Loss: 4.4191
Epoch 1/8... Discriminator Loss: 3.6593... Generator Loss: 7.2007
Epoch 1/8... Discriminator Loss: 0.2579... Generator Loss: 8.8595
Epoch 1/8... Discriminator Loss: 0.0736... Generator Loss: 8.9286
Epoch 1/8... Discriminator Loss: 0.1174... Generator Loss: 3.7036
Epoch 1/8... Discriminator Loss: 0.1500... Generator Loss: 3.0759
Epoch 1/8... Discriminator Loss: 0.0663... Generator Loss: 6.2871
Epoch 1/8... Discriminator Loss: 1.2995... Generator Loss: 0.5007
Epoch 1/8... Discriminator Loss: 0.3359... Generator Loss: 3.9660
Epoch 1/8... Discriminator Loss: 0.0962... Generator Loss: 4.2926

Epoch 1/8... Discriminator Loss: 0.2095... Generator Loss: 2.4974
Epoch 1/8... Discriminator Loss: 0.1690... Generator Loss: 2.8124
Epoch 1/8... Discriminator Loss: 0.2533... Generator Loss: 3.9785
Epoch 1/8... Discriminator Loss: 0.1185... Generator Loss: 3.5219
Epoch 1/8... Discriminator Loss: 0.1402... Generator Loss: 2.8761
Epoch 1/8... Discriminator Loss: 0.0221... Generator Loss: 7.4156
Epoch 1/8... Discriminator Loss: 0.0262... Generator Loss: 10.0951
Epoch 1/8... Discriminator Loss: 0.0753... Generator Loss: 7.8343
Epoch 1/8... Discriminator Loss: 0.2624... Generator Loss: 5.1199
Epoch 1/8... Discriminator Loss: 0.1276... Generator Loss: 2.9867

Epoch 1/8... Discriminator Loss: 0.1473... Generator Loss: 3.4870
Epoch 1/8... Discriminator Loss: 2.0286... Generator Loss: 10.7940
Epoch 1/8... Discriminator Loss: 0.0581... Generator Loss: 5.5782
Epoch 1/8... Discriminator Loss: 0.0292... Generator Loss: 5.9887
Epoch 1/8... Discriminator Loss: 0.0228... Generator Loss: 5.7848
Epoch 1/8... Discriminator Loss: 0.6520... Generator Loss: 4.2834
Epoch 1/8... Discriminator Loss: 0.0856... Generator Loss: 9.0541
Epoch 1/8... Discriminator Loss: 0.2712... Generator Loss: 2.5200
Epoch 1/8... Discriminator Loss: 0.0125... Generator Loss: 12.5157
Epoch 1/8... Discriminator Loss: 0.1376... Generator Loss: 6.4900

Epoch 1/8... Discriminator Loss: 0.1889... Generator Loss: 10.8197
Epoch 1/8... Discriminator Loss: 1.4335... Generator Loss: 0.8737
Epoch 1/8... Discriminator Loss: 0.0880... Generator Loss: 3.6030
Epoch 1/8... Discriminator Loss: 0.0603... Generator Loss: 4.0201
Epoch 1/8... Discriminator Loss: 0.1194... Generator Loss: 3.2700
Epoch 1/8... Discriminator Loss: 0.1757... Generator Loss: 2.6200
Epoch 1/8... Discriminator Loss: 0.1670... Generator Loss: 3.5605
Epoch 1/8... Discriminator Loss: 0.0119... Generator Loss: 8.9914
Epoch 1/8... Discriminator Loss: 0.0153... Generator Loss: 9.1321
Epoch 1/8... Discriminator Loss: 0.0402... Generator Loss: 4.2750

Submitting This Project¶

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.